Diagnosing childhood diseases can be particularly challenging, especially in areas with limited access to specialists. This case report explored the application of MedBrain®, an artificial intelligence–powered Clinical Decision Support System (CDSS), in supporting diagnosis in paediatric practice. We presented two clinical cases from Nigeria: a 4-year-old girl diagnosed with urinary tract infection and a 3-year-old girl diagnosed with acute gastroenteritis. These cases were selected to demonstrate common but diagnostically challenging paediatric conditions with overlapping symptoms. For each case, MedBrain® was used prospectively to generate a ranked list of differential diagnoses with corresponding diagnostic confidence scores based on patient symptoms and clinical findings. The system utilized a hybrid algorithm combining rules-based logic and machine learning to assess input data and compute likelihood estimates. In both cases, MedBrain®'s top-ranked diagnoses— urinary tract infection (96%) and gastroenteritis (95%)—were confirmed by attending paediatricians, validating its clinical utility. Comparatively, the standard clinical diagnosis was initially uncertain in both scenarios due to nonspecific presentations. These findings emphasized the potential of MedBrain® to augment paediatric diagnostic accuracy, particularly in low-resource or non-specialized settings. Future studies should evaluate MedBrain® in larger prospective cohorts or randomized control trials and compare its diagnostic performance with established clinical guidelines.